Solvents are not normally stowed in the engine room because of the danger of accidental spills and because they are a fire hazard. In this case the pouring of a solvent on the side of the hot main engine most likely created greater than normal volume of gases that ignited on the exhaust. Excess gases were also sucked into the turbo-charger intakes spreading the fire to the internals of the engine. The fire spread rapidly into the fibre-board acoustic tiles and wood grounds on the engine room deckhead. The acoustic tiles and wood grounds had 34 years of exposure to both heat and oil mist normally present in an engine room and would require little flame contact to catch fire. The small holes in the tiles allowed the fire to reach the wood grounds behind the tiles. The holes also supplied air to the fire in the grounds allowing the fire to spread across the deckhead behind the tiles. The main engine cannot be stopped if it is running above idle speed. When the skipper first tried to stop the engine, he could reach the stop but not the throttle control at the starboard aft corner of the engine. Consequently the engine continued to run at full speed until the deckhand in the wheelhouse cut it back to idle from the wheelhouse control. At any time after the engine was cut back from the wheelhouse it could easily have been stopped. Release of the CO2 with the main engine running and the vents still open had little or no effect on the fire. Because the main engine was running, most, if not all, of the CO2 was likely carried away through the engine exhaust. The volume percentage for Fixed Gas Smothering Installations in Schedule III of the Regulations does not require the CO2 to completely fill the engine room and casing. When the crew re-entered the engine room, they did not test the atmosphere, yet they suffered no ill effects from the re-entry. This indicates that the concentration of CO2 in the engine room was well below the level it should have attained had the engine room been closed off and the engine stopped when the CO2 was released. The owners of the tug do not supply their vessels with solvent, instead they supply an approved bilge cleaner. There is no record of how or when a can of solvent was left in the engine room. It is likely that the can was left over from work done in the engine room by an outside contractor and was not susequently removed from the space.Analysis Solvents are not normally stowed in the engine room because of the danger of accidental spills and because they are a fire hazard. In this case the pouring of a solvent on the side of the hot main engine most likely created greater than normal volume of gases that ignited on the exhaust. Excess gases were also sucked into the turbo-charger intakes spreading the fire to the internals of the engine. The fire spread rapidly into the fibre-board acoustic tiles and wood grounds on the engine room deckhead. The acoustic tiles and wood grounds had 34 years of exposure to both heat and oil mist normally present in an engine room and would require little flame contact to catch fire. The small holes in the tiles allowed the fire to reach the wood grounds behind the tiles. The holes also supplied air to the fire in the grounds allowing the fire to spread across the deckhead behind the tiles. The main engine cannot be stopped if it is running above idle speed. When the skipper first tried to stop the engine, he could reach the stop but not the throttle control at the starboard aft corner of the engine. Consequently the engine continued to run at full speed until the deckhand in the wheelhouse cut it back to idle from the wheelhouse control. At any time after the engine was cut back from the wheelhouse it could easily have been stopped. Release of the CO2 with the main engine running and the vents still open had little or no effect on the fire. Because the main engine was running, most, if not all, of the CO2 was likely carried away through the engine exhaust. The volume percentage for Fixed Gas Smothering Installations in Schedule III of the Regulations does not require the CO2 to completely fill the engine room and casing. When the crew re-entered the engine room, they did not test the atmosphere, yet they suffered no ill effects from the re-entry. This indicates that the concentration of CO2 in the engine room was well below the level it should have attained had the engine room been closed off and the engine stopped when the CO2 was released. The owners of the tug do not supply their vessels with solvent, instead they supply an approved bilge cleaner. There is no record of how or when a can of solvent was left in the engine room. It is likely that the can was left over from work done in the engine room by an outside contractor and was not susequently removed from the space. Solvent, which had been left in the engine room, was used on the side of the main engine to clean caked-on oil while the engine was running. Solvent fumes were sucked into the air intakes of both starboard side turbo-chargers. Solvent fumes ignited on the hot starboard engine exhaust. The fire spread into the fibre-board acoustic tiles and wooden grounds in the engine room deckhead. Cylinders and pistons were damaged by powder from the dry powder chemical extinguisher used to extinguish the fire in the vicinity of the main engine. Continuing to run the main engine with the engine room vents open circulated air which fed the fire in the deckhead. CO2 from the tug's fixed gas smothering system was released into the engine room without stopping the main engine or shutting off the ventilation. The engine room was re-entered by crew without breathing apparatus, after the release of the CO2, and without checking atmospheric oxygen levels for safe entry. The solvent was not supplied by the vessel's owners and its origins are unknown. The solvent was not supplied by the vessel's owners and its origins are unknown.Findings Solvent, which had been left in the engine room, was used on the side of the main engine to clean caked-on oil while the engine was running. Solvent fumes were sucked into the air intakes of both starboard side turbo-chargers. Solvent fumes ignited on the hot starboard engine exhaust. The fire spread into the fibre-board acoustic tiles and wooden grounds in the engine room deckhead. Cylinders and pistons were damaged by powder from the dry powder chemical extinguisher used to extinguish the fire in the vicinity of the main engine. Continuing to run the main engine with the engine room vents open circulated air which fed the fire in the deckhead. CO2 from the tug's fixed gas smothering system was released into the engine room without stopping the main engine or shutting off the ventilation. The engine room was re-entered by crew without breathing apparatus, after the release of the CO2, and without checking atmospheric oxygen levels for safe entry. The solvent was not supplied by the vessel's owners and its origins are unknown. The solvent was not supplied by the vessel's owners and its origins are unknown. The direct cause of the fire was solvent gases, which were given off by solvent poured onto the side of the running main engine, and which ignited on contact with the hot engine exhaust on the starboard side of the engine. When the grounds and the acoustic tiles covering the engine room deckhead caught fire from the flames on top of the engine, the fire spread behind the tiles, along the grounds making it impossible to extinguish the fire without pulling down the tiles to expose the grounds. Solvent that had been left in the engine room was not removed from the space making it available for use when the skipper was cleaning the engine room.Causes and Contributing Factors The direct cause of the fire was solvent gases, which were given off by solvent poured onto the side of the running main engine, and which ignited on contact with the hot engine exhaust on the starboard side of the engine. When the grounds and the acoustic tiles covering the engine room deckhead caught fire from the flames on top of the engine, the fire spread behind the tiles, along the grounds making it impossible to extinguish the fire without pulling down the tiles to expose the grounds. Solvent that had been left in the engine room was not removed from the space making it available for use when the skipper was cleaning the engine room.